The Agent Skills Directory

[SAFE]: The skill consists exclusively of Markdown documentation and instructional frameworks. No executable scripts (.sh, .py, .js), binaries, or package manifests (package.json, requirements.txt) are present.
[SAFE]: No network operations (curl, wget, fetch) or remote data requests are performed. The skill operates entirely on text provided in the conversation context.
[SAFE]: The reference to reading .claude/product-marketing-context.md is a standard practice for project-specific agents to maintain brand consistency and does not involve exfiltration of sensitive system data.
[SAFE]: There are no attempts at prompt injection, obfuscation, or persistence. The instructions are transparent and aligned with the stated purpose of copy editing.

copy-editing